#Vault Breakers: Super Score — Seeded Simulation Evidence, draft-v0.2

DEV ONLY · NOT FOR PRODUCTION · NOT FOR CERTIFICATION. Every figure here is a MONTE CARLO ESTIMATE. The exact results live in ../exact/draft-v0.2.md and are never re-derived by sampling.

Supersedes draft-v0.1.md, which is retained unchanged.

#Commands

cd /Users/admin/kiro/backend
go test -tags vbdraft -count=1 -v -timeout 180m \
    ./libs/mathengine/games/vault_breakers_super_score -run Draft

#Run A — whole-game confirmation

The master seed string still says v0.1. It is the seed scheme, not a version claim, and it is left alone on purpose: changing it would make v0.1 and v0.2 differ by both the config and the seeds, and only one of those is the thing under test.

#Shard boundaries and seeds

#Results — all estimated

Target 96.00% lies inside the interval, 0.76 standard errors above the estimate.

#RTP decomposition

The free-spins term measured by RunBatch is one number, 12.5570%, split by the regular/charge ratio measured directly on runtime-shaped sequences in Run C (4.55% / 95.45%). The split is derived, the total is measured, and neither is counted twice.

#Win bands

#Exceedance

#Run B — distribution, percentiles and charge behaviour

#Percentiles — all estimated

P50 and P75 are zero because 80.2% of paid rounds pay nothing.

#Base-game charges

Free-spin charges are settled inside the feature and are not itemised here.

Charges in the base game are rare by design: at 30 out of 50,000 they are a visitor to a paid spin, and the mechanic lives in free spins, where the weight is 7,250.

#Charge value distribution — 96,847 drawn

#Feature frequency

#Run C — the free-spins split, retriggers and both buys

Sequences are played the way the runtime plays them: one spin per request against the chosen variant, retriggers resolved from the engine's own verdict via mathengine.ResolveOutcomeTrigger, and the whole-cycle ceiling ending generation.

#Results — all estimated

The regular parts agree to within 0.12% — as they must, since the two modes differ only in their charge pool. That agreement is itself a check that the variant pool override touches nothing else.

#Buy prices — the reference values hold

Both design prices land within half a percent of their fair value. In draft-v0.1 the same prices returned 15.52% and 10.94%; the difference is free spins running their own charge-dense reel table rather than the base one.

#Reconciliation and tolerance

Per-round σ is 13.531314.

whole-game RTP = 95.8973%  +/-0.2652 pp   (95% CI, 100,000,000 rounds)
target         = 96.0000%
deviation      = -0.1027 pp  =  0.76 standard errors
verdict        = target inside the interval

#Sizing note for the next tuning pass

The final corrections were made on the scatter-class weight, because it is the one lever that moves the whole game without moving either buy — a buy is priced off the sequence award, and the sequence award does not depend on how often a sequence is triggered for free.

P(4+ of the class) does not scale as the fourth power of the class weight over this range. Measured across two 100,000,000-round confirmations the effective exponent is about 2.78. Assuming 4 overshot the first correction by a third of a point and cost a run.

#Unresolved

1. The free-spins regular/charge split is derived from Run C's ratio, not measured inside Run A. The atomic executor reports a sequence as one number.
2. Charges collected inside free spins are not itemised; only base-game charges are.
3. A free-spin snapshot pays a regular win on 8.23% of snapshots against the base game's 19.76%, because charges occupy 14.5% of the grid. Free spins are rarer wins, far larger ones, and 95%+ of a sequence's award is the charge part. Deliberate, recorded rather than smoothed, and worth a product decision before this leaves Draft.